For the complete documentation index, see llms.txt. This page is also available as Markdown.
Raster visualization issues at low zoom levels
Raster data looks incorrect at low zoom levels
If your raster data does not render correctly when zooming out, this is usually caused by the interpolation method used during ingestion.
Interpolation determines how pixel values are aggregated when the raster is downsampled for lower zoom levels.
Choosing the correct interpolation method
Nearest neighbor interpolation
Use this when your data represents categorical values or classes.
This method preserves original values and prevents class mixing. It ensures that discrete categories remain visible at all zoom levels and do not “disappear” when the data is generalized.
Linear interpolation
Use this for continuous data, such as:
elevation models
temperature
satellite imagery or other optical data
This method produces smoother transitions between values, reducing pixelation when zooming out.
How to set interpolation
You can define the interpolation method during upload in the Ellipsis Drive interface.
It can also be configured programmatically via the API or using the Python package.
Important note
Choosing the wrong interpolation method can significantly affect how your data is perceived at different zoom levels, especially in derived visualizations or analytical workflows.
